Evaluation of the End of Life Care Train the Trainer (TTT) Education Model

Transcription

1 Evaluation of the End of Life Care Train the Trainer (TTT) Education Model April 2014 Correspondence to: Claire Goodman Professor in Health Care Research University of Hertfordshire CRIPACC Centre for Research in Primary and Community Care College Lane Hatfield, Hertfordshire AL10 9AB Version 2, July 2014, Page 1 of 83

2 Evaluation of the Train the Trainer End-of-Life Care Education Model Research Team Claire Goodman Andrea Mayrhofer Melanie Handley Nigel Smeeton Sarah Amador Sue Davies David Maher Professor in Health Care Research, Centre for Research in Primary and Community Care (CRIPACC), University of Hertfordshire Research Fellow, CRIPACC, University of Hertfordshire Research Assistant, CRIPACC, University of Hertfordshire Social Statistician, CRIPACC, University of Hertfordshire Research Fellow, CRIPACC, University of Hertfordshire Research Fellow, CRIPACC, University of Hertfordshire Senior Lecturer University of Hertfordshire (involved at the design stage of the project) Acknowledgements The University of Hertfordshire would like to extend a very special thank you to the many individuals who participated in this study: Vanessa Convey, Programme Manager, Palliative and End of Life Care Workforce (Education and Training) Health Education East of England, whose support was critical to the success of the project. TTT Project Leads Sarah Russell, Nicki Hallett, Emma Harris and Carolyn Doyle who supported the End-of-Life Care Educators/Facilitators (EFEs) across their regions. Many thanks go to the EFEs for their patience, for caring for their trainers, and for coordinating the flow of information and meetings between care homes, trainers, and researchers. Much appreciation goes to the trainers who, in addition to their varied and pressurised professional roles, took on the role of trainer for this pilot project. Last but not least, many thanks to the care home managers and staff who supported everyone s efforts. This report should be referenced as follows: Goodman C, Mayrhofer A, Handley M, Smeeton N, Amador S, Davies S. (2014). Evaluation of the Train the Trainer (TTT) End-of-Life Care Education Model. (Version 2). Centre for Research in Primary and Community Care (CRIPACC), University of Hertfordshire. Hatfield: University of Hertfordshire. DISCLAIMER: This report presents findings of an independent evaluation commissioned by Health Education East of England. The views expressed in this report are those of the authors and not necessarily those of the Health Education East of England or of any of the care homes involved. Version 2, July 2014, Page 2 of 83

3 TABLE OF CONTENTS Page Research Team 2 Acknowledgements 2 Glossary 7 Executive Summary 8 1 INTRODUCTION AND BACKGROUND Introduction Background to the EoL Care Train the Trainer (TTT) Education Model Evaluation framework Research questions 18 2 METHODS The Train the Trainer (TTT) Pilot Pilot setting Pilot timeline Recruitment Data collection Data collection instruments Data collection instruments for quantitative data Data collection instruments for qualitative data Data analysis Analysis of qualitative data Analysis of quantitative data Interpretation of data Ethical issues: Consent, anonymity and confidentiality 24 3 FINDINGS Care home characteristics Residents characteristics Module delivery: staff availability, blended learning, bite-size teaching Professional roles of EFEs delivering TTT to trainers Professional roles of trainers delivering TTT to learners 37 Synopsis 37 Version 2, July 2014, Page 3 of 83

4 3.6 What evidence is there that the TTT Education Model increases residents, family members, and care home and NHS staff engagement with Advance Care Planning (ACP) and on-going conversations about EoL Care? EoL conversations with residents and families: relational uncertainty Advance Care Planning CP in participating care homes: participants perceptions Review of decedents notes: evidence of Advance Care Planning Place of death Preferred priorities for care (PPC) 43 Synopsis What impact does the TTT model have on CH staff in relation to confidence, reported knowledge, symptom assessment and management, and involvement with NHS services when providing EoL Care? Reported knowledge on symptom assessment and management: pathway uncertainty Comparing past and present care Symptom management as reported by trainers Symptom management: review of decedents notes Formal protocol / Integrated Care Pathway Use of anticipatory medication Certainty and uncertainty of dying pathways Where are residents admitted from and admitted to? Involvement with NHS Services when providing EoL Care: service uncertainty Improved relationships with GPs and HC professionals Feedback from GPs 50 Synopsis Which elements of the programme were the most effective? Barriers and facilitators The Trainer role Care home engagement with TTT The cumulative impact of ABC/TTT/ other EoL Care training Contextual influences: do previous working relationships between EFE and trainer influence the uptake of TTT? Level of EFE support 57 Synopsis Sustainability of the TTT Model Avoiding dilution: the need for EFEs professional input Staff selection 58 Version 2, July 2014, Page 4 of 83

5 4 DISCUSSION Care home and residents characteristics Uptake and implementation of the TTT programme TTT approach The contribution of EFEs Does the TTT model support an increase in quality of EoL Care? The ability of the TTT programme to address uncertainty 65 5 CONCLUSIONS 65 6 LIMITATIONS OF THE STUDY 67 7 RECOMMENDATIONS 67 8 RECOMMENDATION FOR FURTHER RESEARCH 69 REFERENCES 70 APPENDICES Appendix A: Modules of ABC and TTT Education Programmes Appendix B: Resident Service Use Logs Appendix C: Data extraction forms adapted from InterRAI assessment instruments Appendix D: Data extraction forms used for decedent notes Appendix E: Erratum, issued 1 st July 2014 Version 2, July 2014, Page 5 of 83

6 Tables Table 1 EFEs hosting organisations by Site 20 Table 2 Type of CH, type of ownership, and additional training by Site 26 Table 3 Number of beds in each care home by Site (n=17) 27 Table 4 Number of residents followed in each site by number of beds (n=274) 27 Table 5 Admitted from by Site (n=241) 28 Table 6 Admitted from by Care Home Registration Type (n=241) 28 Table 7 Funding by care home registration (n=219) 28 Table 8 Funding by Site (n=219) 29 Table 9 Median length of stay in months by care home (n=234) 29 Table 10 Number of Residents needing extensive assistance in ADLs by Care Home Registration Type 30 Table 11 Number of residents with diagnosis 31 Table 12 Diagnosis of residents by care home registration type 31 Table 13 ACP by cognitive skills for daily decision making (n=224) 32 Table 14 Hospital Admissions in 3 months prior to the start of the study and during 12 weeks of data collection 33 Table 15 EFE contact hours and contact frequency 36 Table 16 Evidence of Advance Care Planning in decedents notes (n=150) 41 Table 17 Evidence of ACP discussion by expectation of death (n=150) 41 Table 18 Evidence of ACP recorded by expectation of death (n=150) 41 Table 19 Evidence of DNACPR by expectation of death (n=150) 41 Table 20 Length of time between DNACPR being signed and death (n=71) 42 Table 21 Place of death by type of care home (n=150) 42 Table 22 Place of death: CH with additional EoL training (n=150) 43 Table 23 Preferred place of care by type of care home (n=92) 43 Table 24 Evidence of PPC by expectation of death (n=150) 43 Table 25 Use of symptom assessment tools by expectation of death (n=150) 47 Table 26 Evidence of protocol / ICP by expectation of death (n=150) 47 Table 27 Use of anticipatory drugs by expectation of death (n=150) 47 Table 28 Learners trained in relation to professional roles of trainers 53 Table 29 Factors that support the sustained implementation of the TTT Programme 64 Table 30 How did the elements of TTT address relational, pathway and service uncertainties? 65 Figures Figure 1 TTT Team configuration between EFEs, trainers and learners 20 Figure 2 Process of data analysis for qualitative data 24 Figure 3 Impact of TTT on ACP and EoL Care in participating care homes 40 Version 2, July 2014, Page 6 of 83

7 Glossary ABC ACP ADL A&E CH CQC DN DNACPR EFE EoE EoL Care GP GSF ICP OOH PPC TTT End of Life Care Education Programme Advance Care Planning Activities of Daily Living Accident and Emergency Care Home Care Quality Commission District Nurse Do not attempt cardio pulmonary resuscitation End of Life Care Educator & Facilitator East of England End of Life Care General Practitioner Gold Standard Framework (EoL Care Education Programme) Integrated Care Pathway GP Out-of-Hours visits Preferred Priorities for Care Train the Trainer EoL Care Education Programme NVIVO SPSS Qualitative Data Analysis Computer Software, QSR V10 IBM Software package for statistical analysis V20 Version 2, July 2014, Page 7 of 83

8 Executive Summary Care home residents have multiple health needs that are often complicated by the presence of dementia. This means that they rely on a range of health and social care staff as well as family members to provide care and make decisions on their behalf. Older people do not enter a care home to die, although they are in the last years of life. For this population it is often unclear, to those providing care, whether a resident is actively dying or experiencing an episode of ill health that they will recover from. End of life care training and support for care home staff can enable staff to incorporate knowledge and skills about anticipating and providing care for residents who are dying into their day to day practice. It aims to ensure that older people can receive care that is appropriate within the care home, is informed by residents wishes about where and how they die, and avoids hospital based interventions that neither prolong life nor provide comfort. Chapter 1: Introduction In October 2012 the Train the Trainer End of Life Care Education Programme (TTT) was commissioned by NHS Health Education East of England (formerly East of England Multi-professional Deanery). The programme built on the success of the ABC End of Life Education Programme that had trained approximately 4000 care home staff across the East of England in End of Life Care (EoL Care). The goal of the TTT pilot project was to consolidate the success of the programme, increase the capacity of the care home workforce to provide end of life care and develop a programme that could sustain training in and provision of End of Life Care in care homes. It used a mix of online resources, skills workshops on teaching and learning, and on-going facilitation from End-of-Life Educators / Facilitators (EFEs) with a professional background as specialist palliative care nurses, community nurses, and acute experience nurses with a passion for EoL Care. Two trainers per care home who had completed the ABC training were invited to join the TTT programme to train six of their peers (Learners) in EoL Care. The pilot ran for nine months (Oct 2012 June 2013), and initially recruited 36 trainers from 18 care homes across the East of England. The evaluation tracked the four stages of the programme and continued data collection up to 4 months after its completion. The organising framework for the evaluation of the pilot was informed by a recognition that uncertainty is a defining characteristic of end of life care for this population and that it is often manifest in three ways, that is, uncertainty about whether the person is dying (pathway uncertainty), uncertainty about how to make decisions about treatments that are in the person s best interests and who should lead that Version 2, July 2014, Page 8 of 83

9 decision making process (relational uncertainty) and finally uncertainty about the ability of the workforce and the visiting services to have the capacity to provide end of life care in the care home (service uncertainty). The evaluation focused on the ability of the TTT programme to address these uncertainties and either resolve them or provide support mechanisms that meant staff were able to hold and manage times of uncertainty. The research questions focused on what supported or hindered the uptake of the programme and specifically on the introduction of Advance Care Planning (ACP), the impact of the programme on staff confidence, knowledge of and skills in end of life care, involvement with the wider systems of care and the perceived costs of the programme. Chapter 2: Methods A mixed method design drawing on qualitative and quantitative methods of data collection was used. This included observation of the 6 workshops in the three sites, baseline assessment of the characteristics and resource use of a randomly selected sample of 274 care home residents who were tracked for 12 weeks, interviews with participating staff, focus groups, review of programme documentation, analysis of trainers audio diaries and a detailed review of the care notes of 150 residents who had died between October 2012 and June 2013, when the pilot programme ended. Seventeen care homes participated, 8 of which had on-site nursing provision. Seven of these had completed or were in the process of completing End of Life Care training additional to the ABC and TTT programme. Thirty Trainers were involved in the pilot and by the end of data collection 114 learners had completed the TTT programme. Two care homes were unable to train any learners during the study period, one needed extensive support to be ready to implement the programme and one had organisational changes that meant they could no longer continue with the programme. Staff who had protected time to undertake the learning and were able to complete the modules as part of their work were in the minority. For those with limited online access and/or those who completed the modules in their own time the involvement of the EFE and opportunities to discuss End of Life Care as part of their work helped to augment their learning and maintain interest in the programme. Chapter 3: Findings - Characteristics of care homes and residents There was a statistically significant difference between care homes with and without on-site nursing, with the latter having fewer hospital admissions. Version 2, July 2014, Page 9 of 83

10 Seventy nine per cent of the residents who died had died in the care home, a finding that is consistent with the earlier evaluation of the ABC programme. Just over half of the randomly selected 274 residents had an advance care plan (ACP) in place and the majority (95%) did not have an unplanned admission during the time of data collection. The review of decedents notes found evidence of discussions relating to ACP for 111 residents (74%) and detailed discussions of symptom assessment and management. Most activity related to ACP and EoL Care was concentrated in the last week and days of life. Over a quarter of those who died had a DNACPR in place the week before death. Of 92 residents, 80 (87%) died in their preferred place of death, which included one person dying in hospital as specified, and one person where care was appropriate for their needs. Of the 118 residents whose death was expected 63% has anticipatory medication in place. An unexpected finding was that of the 129 residents who had a recorded reason for admission, 23 (18%) had been admitted from hospital specifically with a diagnosis of dying and for end of life care. This would suggest that some of the care homes were being recognised for their expertise in EoL Care. Nearly two-thirds (65.7%) of residents had a diagnosis of dementia. Benefits of the programme The communication module and the ACP module were fundamental for trainers and learners. The flexibility and blended learning approach of the TTT programme, which combined the use of online learning and face to face and practice based teaching with expert facilitation, were seen as strengths of the programme. For care homes that had experience of using other EoL resources the context sensitive approach of the TTT programme was identified as preferable and more person centred. Previously, nine times out of ten we d just go oh, if a resident Version 2, July 2014, Page 10 of 83

11 approached this subject. But now we talk about it it makes one aware that one can talk about EoL one can assist relatives to face what s coming [FG0106P3]. The majority of care home staff believed that the TTT programme had given them confidence in providing EoL Care, encouraged the initiation of conversations about residents priorities and preferences for EoL Care, and helped in how they worked with visiting GPs in planning care. we used ACP check-lists before the training, but they have been amended since the training. In addition, we now use original DNACPR forms, whereas before we just noted it in the care plans; now we have a file of the originals [T02111SLD]. There were also numerous examples given of when they would provide anticipatory care and be proactive in the assessment and management of symptoms of pain and distress. There was limited evidence of the programme changing how the care homes worked with wider NHS services and whether it increased their access to specialist support. The availability and access to an EFE was crucial, particularly in situations where there was limited online access, and when participating care homes had limited prior experience of providing EoL Care or of training learners in the work place. Care home readiness Care home readiness was a consistent finding that was threaded through the accounts of participants as a key influence on uptake and successful completion of the programme. For some care homes a prior history of working with end of life care resources facilitated relationships with service providers such as GPs and hospitals, either in relation to prescribing anticipatory medication, issuing signatures for DNACPR forms, or referring patients to the care home for palliative care. Better service relationships were also influenced by pro-active and clinically qualified managers. In combination, such factors enhanced the implementation of the TTT EoL Care education programme. This is to be contrasted to care homes that were in flux during the TTT pilot programme, either due to managerial change, staff turnover, or conflicting priorities at the time. Version 2, July 2014, Page 11 of 83

12 Variation Variation in the professional roles and responsibilities of the trainers who taught the learners also affected uptake and implementation. Trainers who had the most opportunity to teach learners were those whose roles allowed them to spend time on the floor with learners and offer bite-size, applied teaching as and when opportunities arose. To be able to act on their learning trainers and learners also needed to have a level of authority and responsibility. Across the 17 care homes the evaluation found a wide variation of care home experience and prior knowledge of End of Life Care. The blended learning approach of the TTT programme meant that it was able to respond and adapt to the different levels of care home readiness. It was a programme that was valued by the care home staff, and participating care homes were able to sustain patterns of care observed from the earlier ABC evaluation (Pyper et al 2013). Contributions There are two particular contributions of the TTT programme. Firstly, to build on learning from the ABC model and extend the involvement of care home staff who otherwise would not have access to End of Life Care training. Secondly, to augment, reinforce and consolidate learning either from the use of other EoL resources or the expertise of key practitioners in the care home. The TTT programme was preferable to paper based initiatives and the care homes valued the flexibility that the programme offered. Perceptions The TTT EoL Care training was not perceived as a course that had to be passed, but as an approach that took into consideration the vulnerabilities, sensibilities and constraints that are part of the home contexts that EoL Care training is meant to strengthen. It needed the EFEs to make that happen and relied heavily on their commitment as well as on the goodwill and interest of the staff to undertake work in their own time. Where the latter were not in place there were few incentives to participate, or sanctions that could be used. Version 2, July 2014, Page 12 of 83

13 Managerial support Managerial support was critical in three ways, in making sure that staff would be available, that they would have protected time, and in encouraging learners to apply their learning when working with colleagues and NHS professionals. Care homes without on-site nursing needed more support and facilitation, and future implementation may want to target these care homes and prioritise those care homes that have had less access to End of Life Care resources. The evaluation of the TTT programme demonstrated the importance of a long term commitment to working with care homes, and the value of a programme that can reinforce and sustain learning in organisations where staff and management often change, and where residents dying trajectories are protracted and complicated by the presence of dementia and multiple morbidities. Recommendations The TTT programme with its blended learning approach should be continued and extended to complement the achievements of the ABC programme, in order to sustain the development of knowledge and skills and the implementation of effective End of Life Care in care homes. Care homes without on-site nursing should be targeted ahead of care homes with on-site nursing. The role of the facilitator is critical to achieving this. The EFE s role should be one that is based on a qualification in palliative care and ongoing experience in EoL care. A careful pre-assessment and discussion with the care home could inform a negotiated agreement about frequency and availability of EFE support with the recognition that some care homes will need more support than others. The findings suggest that, whilst EFE support can reduce as trainers and learners grow in confidence and skills, there is an ongoing need for EFE review and support, particularly at times of organisational change or turbulence in the care home. A review of EFE approaches to providing support should consider what a minimum level of support for a participating care home would require in EFE time, for different levels of care home readiness and whether there is an on-site clinician or not. Version 2, July 2014, Page 13 of 83

14 The costing of EFE support should be calibrated to reflect the readiness of the care home and care home staff to participate in the programme. It also needs to take account of EFE time alongside use of other NHS services, especially in cases where there is some blurring between an EFE s role as TTT facilitator, and their presence in the care home in their role of clinical nurse specialist. Staff selection, both for trainers and learners, is a critical consideration for the TTT model. Aspects such as prior experience, opportunity, authority and responsibility should be discussed, assessed and monitored. The commitment of management, priorities and interim goals of the participating care home staff, and internet availability should be discussed prior to embarking on a TTT programme as components of care home readiness to participate. The development and testing of a care home readiness assessment tool to inform recruitment of care homes and levels of EFE support provided to care homes. Care homes should consider organising practice based teaching in units within the care home. Version 2, July 2014, Page 14 of 83

15 1 INTRODUCTION AND BACKGROUND 1.1 Introduction This report presents the findings from the evaluation of the Train the Trainer (TTT) programme. It provides a brief introduction and background to the TTT pilot programme in the East of England and summarises how the evaluation was approached. Findings are presented in two sections. The first section discusses information about the participating care homes, resident characteristics, use of services and related End of Life Care outcomes. The second section provides a description of how the programme was implemented, of participants views and experiences and what supported and inhibited its uptake and implementation. The report ends with a discussion of the impact of the programme and makes recommendations for future implementation and research. 1.2 Background to the Train the Trainer Education Model In the UK the majority of long term care for older people is provided by independent care providers. Care home staff are experienced in caring for frail older people. However, they rely on primary health care for medical support, access to secondary care and generic specialist palliative care. Historically, health care services have an erratic relationship with care homes (Davies, Goodman et al 2011). Structured End of Life Care initiatives such as the Gold Standard Framework for Care homes, and locally the ABC initiative have demonstrated promising results (Gandy, Roe, McClelland & Ashton 2011; Pyper, Sawyer and Pyper & Mayhew, 2013). The challenge is how to sustain these improvements so that care home staff act on the learning acquired, develop their skills and support colleagues in providing end of life care. There is a need to embed end of life care into the day to day work of care homes. The Train the Trainer Education Model (TTT) aims to build on earlier work and offers an approach that retains the expert input and support from specialist palliative care mediated through the targeted development of learning champions. This is the focus of this report. The ABC End of Life Education Programme consisted of the use of online training to deliver six training modules and provide facilitated support from EFEs who held Version 2, July 2014, Page 15 of 83

16 various roles in specialist palliative care nursing, community nursing and acute experience nursing to approximately 4000 care home staff across the East of England. As a result of the success of the ABC training programme (Pyper et al, 2013 [aka PHAST Report]), the Train the Trainer End of Life Care Education Programme (TTT) was commissioned by NHS Health Education East of England (formerly East of England Multi-professional Deanery) to consolidate the learning from the ABC programme and equip more staff to provide End of Life Care. The TTT project aimed to train two trainers per care home who had participated in the ABC programme to support six learners in their care home. Details of the ABC and the Train the Trainer End of Life Care Education Programme are listed on Appendix A. The TTT programme, in addition to the use of the online training from the ABC EoL Care Education Programme for new learners, sought to equip trainers with the skills to disseminate their learning more widely within care homes. Trainers responsibilities therefore included the preparation of on-line and face-to-face teaching sessions, discussions, and whenever possible offering learners bite-size micro-teach sessions in daily practice. Full teaching sessions were observed and evaluated by End of Life Care Educators / Facilitators (EFEs). The design of the TTT pilot study is described in Section Evaluation framework Older people do not enter a care home to die and for this population it is often unclear whether a resident is actively dying or experiencing an episode of ill health they will recover from (Goodman et al 2010). Because of the presence of dementia, there is often uncertainty about residents wishes, compounded by uncertainty about who is ultimately responsible for decisions about how and where EoL Care should be provided, and how palliative care support or services will be provided to care homes if needed. Previous research undertaken by the team identified three types of uncertainty that shape how decisions are made to provide end of life care and, importantly, inform how care is provided and care home staff s ability to support people to die in the care home. These uncertainties relate to the individual s trajectory to death, the relationships that exist between residents, staff, family and visiting health professionals, and to the ability of staff and the capacity of services to provide Version 2, July 2014, Page 16 of 83

17 support when needed. This evaluation structured its data collection and analysis based on the extent to which the TTT education model was able to ameliorate uncertainty in these three areas. Pathway uncertainty The minority of residents in care homes die from cancer. Their pathway to death is often one of progressive deterioration and gradual functional loss. This is sometimes only apparent in retrospect (Barclay et al (in press); Handley et al 2013). The pathway to death may be characterised by incremental changes in function and health, sudden episodes of ill health from which there is a sustained recovery of over six months, a recognisable period of dying or sudden death (Goodman et al 2013). It is very difficult to differentiate between when someone is actively dying and when, for example, it is sensible to refer for a specialist opinion. There can be a fine line between ageism that says it is not worth actively treating older people and recognising when palliative care is the appropriate treatment of choice. Equally, when an older person has dementia and/or a range of health care problems and functional limitations, assessment and treatment of symptoms such as pain, breathlessness and fatigue can be more complicated. End of life care interventions need to be able to equip practitioners with the resources, tools and skills that can address the wide range of diagnostic/assessment and treatment issues that do not reflect cancer trajectories. Relational uncertainty In care homes effective end of life care is shaped by how NHS services work with care homes and manage the division between public and private provision. Often there are multiple professionals, family members and care home staff involved in providing and discussing a resident s care. The difficulties of maintaining communication and consistency between all those involved in a resident s care, concerns about personal and professional liability, and/or weak working relationships can influence decisions to call emergency services, persist in treatments that do not promote recovery and delay conversations about how to plan EoL Care. Interventions that improve and sustain relational working and shared decision making in EoL Care are more likely to support conversations over time and evidence informed clinical practice. Version 2, July 2014, Page 17 of 83

18 Service uncertainty Linked to the issues about bridging the divide between the NHS and long term care providers is uncertainty within the service that the workforce has the capacity to provide appropriate support and that there are resources to provide end of life care. Effective end of life care interventions should therefore be able to demonstrate how they have increased the knowledge and confidence of participants, built sustainability into the service (that addresses the known problem of workforce turnover and variable access to NHS services), and provided access to on-going specialist support and advice. They also need to address the capacity of primary care and linked specialist services to provide the resources (for example equipment) and support residents alongside their existing community dwelling caseload. It is reasonable to assume that improved knowledge about EoL Care and use of certain end of life care tools could address pathway uncertainty and service uncertainty, or that changing patterns of working and how care is structured could help to ameliorate relational uncertainty and embed new practices into every day working. This evaluation framework captures the context, process and outcomes of end of life care Research questions 1) What evidence is there that the TTT Education Model increases residents family members, CH and NHS staff engagement with Advance Care Planning (ACP) and on-going conversations about EoL Care? 2) What impact does the TTT Education Model have on CH staff in relation to: Confidence Reported knowledge Symptom assessment and management Involvement with NHS services when providing EoL Care 3) What elements of the EoL Care Programme were the most effective? 4) What were the barriers and facilitators to the implementation of the programme? 5) What evidence is there that the programme supports an increase in quality of care for older people in care homes identified as in need of end of life care? Version 2, July 2014, Page 18 of 83

19 6) What impact is the programme making on wider systems of care and primary health care in particular? These questions are debated in relation to the evaluation framework. Version 2, July 2014, Page 19 of 83

20 2 METHODS 2.1 The Train the Trainer (TTT) Pilot The Train the Trainer (TTT) education model is a continuation of the ABC EoL Care education programme and was designed to disseminate palliative and EoL Care training throughout East of England (EoE) care homes. The TTT Pilot was planned for 18 care homes. Selected EoL Care experts in Anglia, Hertfordshire and Essex took on the role of EoL Care Facilitators/Educators (EFEs) and supported trainers to deliver the ABC programme throughout the duration of the pilot. As shown in Figure 1, the EFEs are at the core, teaching trainers who in turn train learners. The diagram depicts an increasingly wider range of dissemination of knowledge and learning through the integrated relationships which underpin the design of the TTT Model. Figure 1: TTT Team configuration between EFEs, trainers and learners EFEs held various clinical and education roles and were employed by a range of organisations. Professional roles included palliative link nurse, EoL specialist, hospice nurse, EoL Care Educator, practice development nurse for care homes, palliative nursing and district nursing. Table 1 shows the mix of organisations that hosted the EFEs during the TTT pilot project. Table 1: EFEs hosting organisations by Site Site Site 1 Site 2 Site 3 Hosting organisation NHS Trust, Cancer Network, Hospice Hospices NHS Trusts Version 2, July 2014, Page 20 of 83

21 2.1.1 Pilot setting The pilot was to take place in 18 care homes across the East of England, including: 6 care homes in Anglia (12 trainers and up to 72 learners) 6 care homes in Hertfordshire (12 trainers and up to 72 learners) 6 care homes in Essex (12 trainers and up to 72 learners) Programme terminology used in this report is highlighted below: End of Life Care Educator/Facilitators (EFEs) End of life Care experts who supported care home staff (Trainers) to deliver EoL life care and palliative care training as part of the TTT pilot programme. Trainers Care home staff who had completed the ABC training and had been nominated to deliver EoL Care and palliative care training to their colleagues (Learners) as part of the TTT pilot programme. Learners Care home staff who received EoL care and palliative care training from their colleagues (Trainers) as part of the TTT pilot programme Pilot timeline The pilot programme was delivered in four stages: First Stage: October 2012 Trainers attended an initial 2 day TTT workshop (Workshop 1) in October 2012 at which they were further familiarised with the EoL Care ABC programme, were given guidance around educational and facilitation skills, and had the opportunity to practice the delivery of a training session in the safety of a supportive peer group. Second Stage: November 2012/December 2012 Trainers were expected to start delivering the ABC programme within 4 weeks of attending the TTT workshop. Third Stage: January 2013 Trainers reconvened for a final workshop (Workshop 2) to provide feedback to EFEs, project leads and to each other, and discussed any outstanding issues. Fourth Stage: March/April 2013 Post evaluation to take place no later than 3 months after the TTT programme has completed within the care home. 2.2 Recruitment Care homes, care home managers, EoL Care Facilitator Educators (EFEs), Trainers and Learners were recruited through Health Education East of England. Two care Version 2, July 2014, Page 21 of 83

22 home providers organisations identified care homes that could be invited to participate. Inclusion was determined by the presence of staff that had completed the ABC programme. The intervention was delivered without charge to the care homes. It was made explicit by the funders at the outset of the programme of the scheme that participation in the TTT pilot would involve an independent evaluation. Information leaflets and consent forms about the evaluation were sent out to care home managers, EFEs, trainers and learners, and all participants were offered the opportunity to refuse participation in the evaluation. 2.3 Data collection A mixed methods approach was used, using quantitative and qualitative data collection methods. The evaluation was based on the assumption that care home staff from participating care homes could commit to face to face interviews and the completion of audio diaries, and that care home managers would facilitate the evaluation and linked data collection Data collection instruments To minimise burdening participants with extra work, a light touch approach was proposed, which aimed to maximise learning from routinely collected data, or from data completed as part of the TTT intervention To establish the characteristics of residents and their service use in participating care homes, Resident Service Use Logs 1 and data extraction forms based on InterRAI Assessment instruments 2 were used to gather information from a randomly selected sample of 30% of residents per participating care home. Resident Service Use Logs recorded primary and secondary service use such as GP visits, out of hours (OOH) GP call outs, visits by District nurses (DN), services rendered by allied health professionals, ambulance services, visits to Accident and Emergency (A&E), and unplanned hospital admissions. The modified InterRAI Assessment Instruments were used to elicit residents levels of frailty and dependency, cognition, and overall health condition. Service Use Logs were completed by care home staff in their role of TTT trainers. Data extraction forms were completed by a mix of TTT trainers, EFEs and members of the research team. The 30% sample was followed for 12 weeks. 1 Attached as Appendix B 2 Attached as Appendix C Version 2, July 2014, Page 22 of 83

23 In addition, the research team looked through 150 residents decedent notes, dated from October 2012 to the end of June 2013, from which information pertaining to Advance Care Planning (ACP), preferred and actual place of death, service use, and EoL pathway was extracted 3. Data collected relating to Advance Care Planning in participating care homes captured information of ACP, DNACPR forms completed and signed, place of death, and Preferred Priorities for care (PPC) as recorded in care home notes. Quantitative data were entered into IBM SPSS (Version 20.0). Section discusses data analysis in detail Data collection instruments used for qualitative data included audio diaries, semi-structured interviews with 27 trainers, a focus group with trainers/learners per site, face to face interviews with 10 EFEs, and face to face interviews with two care home managers. Interviews were recorded, transcribed, anonymised, and entered into QSR NVivo 4 (Version 10). The analysis of data is discussed in detail in Section Audio Diaries It was not possible to observe care, and in attempts to involve care home staff as co-researchers in the evaluation, trainers were provided with Kindle Fire tablets and asked to record their reflections on aspects the TTT model. This might have included elements of module delivery, key events in the care homes, reflections on how practice might have changed or remained unaffected by the TTT training, or reflections on conversations with colleagues and/or residents concerning dying and EoL Care. The Kindle Fire tablets were password protected and registered with e- mail accounts linked to the University of Hertfordshire, so that sound files could be ed from the Kindle Fire tablet to the researcher s desk anonymously. The use of the Kindle Fire tablets was demonstrated at the 2 nd workshops. A protocol for their use and an Aide Memoir and were developed for trainers. 2.4 Data Analysis Analysis of qualitative data Qualitative data were collected via semi-structured interviews, focus groups and audio recordings. QRS NVivo (Version 10) was used to support data management. Processes of data analysis are depicted in Figure 2 below: 3 Attached as Appendix D 4 Computer software for qualitative data management Version 2, July 2014, Page 23 of 83

24 Data entry Sorting and structuring information into descriptive categories Thematic Analysis Cross-sectional analysis Verification of findings Analysis of emerging themes in relation to research questions and the conceptual framework of the evaluation Interrogating data across cases and different groups of participants, taking into consideration care home characteristics and contexts Using quantitative data (quality indicators) and explanatory variables to verify findings Synthesis of findings Report writing Drawing together findings based on qualitative and quantitative data analysis; interpretation of findings Presentation of findings, discussion, conclusions and recommendations Figure 2: Process of data analysis for qualitative data Analysis of quantitative data Descriptive statistics derived using SPSS (Version 20) and Stata (Version 11.2) were used to report resident and care home characteristics, estimated service use and costs, and involvement of EFEs. The chi-squared test and Fisher s exact test (2-sided) were used to detect differences between care homes, resident characteristics, patterns of service delivery, and care homes uptake of the TTT programme. 2.5 Interpretation of data The organisation and analysis of data proposed to use the uncertainty framework as set out in Section to evaluate the impact of the model on different aspects of end of life care. The interpretation and discussion of findings is presented in Section Ethical issues: Consent, anonymity and confidentiality Participants were given at least two weeks to decide whether or not to take part in this evaluation. At each part of the process the researcher reiterated that Version 2, July 2014, Page 24 of 83

25 participation was voluntary, and that participants could withdraw at any stage without giving a reason. Any information that may identify participants or care homes was removed in the transcripts. The Research Ethics Committee (National Institute for Social Care and Health Research, Rec Ref 12/WA/0384), reviewed the protocol and study documents and supported the application. Governance approval was obtained from five participating local authorities. Version 2, July 2014, Page 25 of 83

26 3 FINDINGS This section presents an overview of findings pertaining to participating care homes, residents, EFEs and trainers. Care home characteristics are presented in Section 3.1 in relation to care home type, ownership, additional EoL Care training and size. Section 3.2 covers residents characteristics such as admission, funding, length of stay, diagnoses, care needs, and service use. Findings concerning TTT Module delivery, professional roles of EFEs delivering TTT, and professional roles of trainers delivering TTT to learners are discussed in Sections 3.3 to 3.5. Specific research questions pertaining to Advance Care Planning, impact of TTT on care home staff, impact of programme elements, barriers and facilitators, and sustainability are addressed in Sections 3.6 to Care home characteristics The pilot project was designed to include 18 care homes across three regions. Three of these care homes withdrew at the early stages of the project. Two of them were replaced by care homes that were known to EFEs. Seventeen care homes participated in the pilot. The participating care homes were either registered as residential care homes or care homes with on-site nursing and had a mix of ownership. Some care homes had received additional EoL Care training, either prior or during the pilot project (see Table 2). The effect of TTT and additional EoL Care training is discussed in Section Table 2: Type of CH, type of ownership, and additional training by Site Site 1 Site 2 Site 3 Total Care home residential Care home with nursing For profit Not for profit Additional EoL Care training (GSF) completed or in progress Care home sizes ranged from under 40 beds to over 120 beds. The average size of care homes in England and Wales is 36 beds (Laing & Buisson 2009). With the exception of one care home, all care homes in our sample were larger than average (Table 3 refers). This is indicative of a trend for major providers of residential care homes and care homes with on-site nursing to operate larger homes (Laing & Buisson 2009). Version 2, July 2014, Page 26 of 83

27 Table 3: Number of beds in each care home by Site (n=17) plus Total Site Site Site Total Resident characteristics A randomly selected 274 residents from across three sites (30% of the population) were followed for 12 weeks to gain an overview of their characteristics and of their use of primary and secondary care services (See Table 4). Table 4: Number of residents followed in each site by number of beds (n=274) plus Sample Total Population Size % of Population in Sample Site Site * 0 18* Site * Total * Following discussions with staff who would be responsible for data collection it became apparent that, in some cases, staff would be restricted to data collection in their own units. In these instances the number of beds in these units was randomised to produce a 30% sample of the unit population. Sex of Residents Nearly three quarters (74.4%) of the randomly selected sample were female. This reflects national figures (Lievesley, Crosby & Bowman, 2011). Age at Admission The median age of residents at time of admission was 83 years old. Admitted From Information on source of admission was recorded in 241 of the 274 residents. As indicated in Table 5, residents were admitted from home, hospital, other care homes, and other, which included psychiatric wards or institutions for people with learning disability. Version 2, July 2014, Page 27 of 83

28 Table 5: Admitted from by Site (n=241) Site Home Hospital Other Care Other* Total Home 1 39 (35.8%) 24 (22%) 27 (24.8%) 19 (17.4%) (44.4%) 15 (20.8%) 19 (26.4%) 6 (8.3%) (28.3%) 25 (41.7%) 16 (26.7%) 2 (3.3%) 60 Total 88 (36.5%) 64 (26.6%) 62 (25.7%) 27 (11.2%) 241 * Includes categories such as mental health residence, psychiatric hospital or unit, settings for persons with learning difficulties, rehabilitation hospital/unit, sheltered housing, correctional facility, and not known / not recorded Residents admitted from hospital or from another care home were more likely to be admitted to care homes with on-site nursing (p=0.013, chi-squared test) (Table 6 refers). Table 6: Admitted from by Care Home Registration Type (n=241) Care Home Type Home (%) Hospital (%) Other Care Other (%) Total Home (%) Residential 48 (48.5) 23 (23.2) 20 (20.2) 8 (8.1) 99 On-site Nursing 40 (28.2) 41 (28.9) 42 (29.6) 19 (13.4) 142 Total 88 (36.5) 64 (26.6) 62 (25.7) 27 (11.2) 241 Method of Funding Residents who are self-funding are known to have lower dependency on admission and live longer than residents in care homes that rely on public funding. Information on the method of funding was collected for 219 of the 274 residents. As indicated in Table 7, there are more self-funding residents in care homes registered as residential homes than in those with on-site nursing (43.7% v 18.1%). Table 7: Funding by care home registration (n=219) Care Home Registration Social Services Residential 44 (42.7%) With nursing 40 (34.5%) Total 84 (38.4%) NHS 7 (6.8%) 46 (39.7%) 53 (24.2%) Self Funding 45 (43.7%) 21 (18.1%) 66 (30.1%) Unclear 6 (5.8%) 2 (1.7%) 8 (3.7%) SS & private Total top-up and/or NHS and private top-up 1 (1%) (6%) (3.7%) 219 Nearly a quarter (24.2%) of residents was reported as in receipt of NHS funding for their care home placement, which appears high. One care home organisation had a statistically significantly higher proportion of residents in receipt of NHS funding than other care home organisations, which may suggest that these residents had Version 2, July 2014, Page 28 of 83

29 higher care needs. As shown in Table 8, in Site I this is considerably higher at 42%. These residents were concentrated in nursing homes (39.7% versus 6.8% in residential homes). Table 8 below breaks down funding type by care home registration and site. Table 8: Site Funding by Site (n=219) Social Services NHS Self- Funding Unclear SS & private top-up and/or NHS and private top-up 1 24 (27.3%) 37 (42%) 27 (30.7%) 0 (0%) 0 (0%) (54.9%) 5 (7%) 20 (28.2%) 6 (8.5%) 1 (1.4%) (35%) 11 (18.3%) 19 (31.7%) 2 (3.3%) 7 (11.7%) 60 Total 84 (38.4%) 53 (24.2%) 66 (30.1%) 8 (3.7%) 8 (3.7%) 219 Total However, these figures should be interpreted with caution as it was not possible to verify them. Length of Stay At the time of data collection the median length of stay across for residents was 22.6 months ( ). Table 9: Median length of stay in months by care home (n=234) Care Home Median number of months N (range) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) 12 Total 22.6 ( ) 235 Version 2, July 2014, Page 29 of 83